基于损失分布法的我国商业银行操作风险计量的实证研究

发布时间：2018-06-13 17:16

本文选题：操作风险 + 损失分布法　；参考：《江西财经大学》2015年硕士论文

【摘要】：操作风险是商业银行面临最原始的风险,但伴随商业银行发展,操作风险并没有像其他风险一样受到重视,认为操作风险带来的损失是可以忽略的。然而从20世纪80年代起,国外金融机构连续发生了巨额操作风险损失事件,损失事件发生的原因各不相同,加强对操作风险计量和监督已成为商业银行正常经营必不可少的前提。在国内,商业银行对操作风险的认识非常浅显,对操作风险监管缺少相应的法律法规,且没有收集损失数据建立损失数据库,这给我国商业银行操作风险计量带来了巨大的困难。因此,本文在数据缺失的前提下对我国商业银行操作风险进行计量具有重大意义。本文首先介绍了操作风险的背景、研究意义以及国内外相关研究。其次介绍了商业银行操作风险的定义、分类、特点和计量操作风险常用的方法,并比较了这些方法的优缺点,得出损失分布法是现阶段最适合国内商业银行操作风险计量的方法的结论。再次研究了从时间、金额和类型三方面对损失数据进行分析,并介绍了蒙特卡洛模拟下损失分布法的原理和过程。对此,本文以公开发布的1994—2014年国内商业银行操作风险损失数据为样本,运用损失分布模型对国内商业银行操作风险资本金进行估算。在实证过程中,首先对损失强度数据和损失频率数据进行拟合优度检验,拟合出它们的最佳分布。然后运用财务风险管理软件水晶球进行蒙特卡洛模拟,拟合出总体损失分布函数,并通过总体损失分布函数计算出商业银行应该计提的风险资本金。最后对损失强度数据进行对数化处理,拟合出对数化后的损失强度分布,并运用蒙特卡洛模拟方法,计算出改进后的风险资本金。经过改进后,商业银行2015年应计提的操作风险资本金为1300.037亿元,大大低于改进前所需的风险资本金。最后,基于前面的结论,从损失分布法推广和操作风险控制两个方面来对操作风险管理提出建议。这有助于商业银行提高操作风险防范意识,降低操作风险带来的损失。
[Abstract]:Operational risk is the most primitive risk faced by commercial banks, but with the development of commercial banks, operational risk is not paid as much attention as other risks, and the losses caused by operational risk can be ignored. However, since the 1980s, foreign financial institutions have suffered a series of huge operational risk losses, the causes of which are different. Strengthening the operational risk measurement and supervision has become an essential prerequisite for the normal operation of commercial banks. In China, commercial banks have a very simple understanding of operational risk, lack of corresponding laws and regulations on operational risk supervision, and do not collect loss data to establish a loss database. This brings enormous difficulties to the operational risk measurement of commercial banks in our country. Therefore, it is of great significance to measure the operational risk of commercial banks in our country on the premise of missing data. This paper first introduces the background of operational risk, research significance and related research at home and abroad. Secondly, it introduces the definition, classification, characteristics and common methods of measuring operational risk of commercial banks, and compares the advantages and disadvantages of these methods. It is concluded that the loss distribution method is the most suitable method for the operational risk measurement of domestic commercial banks at present. The loss data are analyzed from three aspects of time, amount and type, and the principle and process of the loss distribution method under Monte Carlo simulation are introduced. In this paper, the operational risk loss data of domestic commercial banks from 1994 to 2014 are taken as samples, and the operational risk capital of domestic commercial banks is estimated by using the loss distribution model. In the process of demonstration, firstly, the data of loss intensity and frequency are tested by the goodness of fit test, and their best distribution is fitted out. Then Monte Carlo simulation is carried out by using the financial risk management software crystal ball, and the total loss distribution function is fitted, and the risk capital should be calculated by the total loss distribution function. Finally, the logarithmic data of loss intensity is processed, the distribution of loss intensity is fitted out, and the modified risk capital is calculated by Monte Carlo simulation method. After the improvement, commercial banks should charge operating risk capital of 130.0037 billion yuan in 2015, which is much lower than the risk capital required before the improvement. Finally, based on the previous conclusions, some suggestions on operational risk management are put forward from the aspects of loss distribution method extension and operational risk control. This will help commercial banks to improve the awareness of operational risk prevention and reduce the losses caused by operational risks.
【学位授予单位】：江西财经大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：F832.33

【相似文献】